For indirect proxy datasets we consider the following variables: Sea Surface Temperature (SST), Sea Surface Salinity (SSS), and ocean currents.

1.0 Sea Surface Temperature (SST)

The main bulk of SST data is based on satellite remote sensing, with a limited amount of in situ data.

1.1 Satellite SST sensors

Satellite SST sensors operate either in the infrared (IR) or microwave segment of the electromagnetic spectrum.

AVHRR (Advanced Very High Resolution Radiometer) is an IR sensing instrument in operation since 1979. It is flown on a polar orbit providing almost global coverage twice a day. The resulting SST products are commonly remapped on 1, 4 or 9 km lat/lon grid with various temporal frequencies (daily, weekly or monthly). The advantage of AVHRR data is in their high spatial and temporal resolution, however  they are impacted by clouds leaving large gaps in coverage.

Source of the data:

MODIS (Moderate Resolution Imaging Spectroradiometer) is an IR instrument flown on Terra and Aqua polar orbiting satellites. It is in operation since 1999. The SST products are provided on daily, weekly or monthly resolution and mapped on 4 km lat/lon grid globally. Since MODIS is an IR instrument, it is affected by clouds, which is its main limitation.

Source of the data:

GOES (Geostationary Operational Environmental Satellites) SST product is also based on radiance in IR channels, and thus is also affected by clouds. Since GOES satellite is geostationary the temporal frequency of the data is 3 hours,  but it can be up to 1 hour for selected areas.  The grid spacing of SST product is 4 km at nadir.  The GOES SST data are available since 2009 to present. The spatial coverage is determined by a particular GOES satellite (GOES East at 75.2° W nadir and GOES West at 137.2°W nadir)

Source of the data:

TRMM (Tropical Rainfall Measuring Mission) was in operation from 1997 to 2015. Its Microwave Imager (TMI) instrument is a passive microwave sensor used to derive SST. The main advantage of TMI over IR-based instruments is that it can sense the surface radiance through clouds. However, the spatial resolution is coarser and TRMM/TMI SST products are mapped on ¼ degree lat/lon grid.

The data coverage is from 40S to 40N latitude, with temporal frequencies of daily, weekly or monthly.

Source of the data:

AMSR-2/AMSR-E/AMSR-J  (Advanced Microwave Scanning Radiometer) family of instruments is in operation from 2002 till present. The individual AMSR instruments are identical, but were deployed on different polar orbiting satellites with very similar orbital parameters.  The resulting SST products are available as daily, weekly or monthly global maps on ¼ degree lat/lon grid. Similarly as for TRMM/TMI the AMSR  instrument is not affected by clouds.

Source of the data:

1.2 Satellite SST datasets

Special SST datasets combining several data sources are provided. This approach overcomes some of the sensor’s limitations, e.g. cloud effect for IR. Examples are given below.

OISST (Optimum Interpolation Sea Surface Temperature – also known as Reynolds SST) is SST product which combines various satellite and in situ (buoy) data.   The interpolation algorithm fills in gaps due to instrument limitations (e.g. clouds) while keeping higher spatial resolution. The OISST data start from 1981 up to present.  The coverage is global on ¼ deg. lat/lon grid with daily frequency.

Source of the data:

GHRSST (Group for High Resolution Sea Surface Temperature) dataset is high resolution SST product, which utilizes IR (AVHRR. MODIS), microwave (AMSR) and in situ (buoys) SST observations.  The data span is from  May 2002 to present. The SST data are produced daily on 0.01  deg.  lat/lon  global grid.

Source of the data:


Sentinel-3 STLSTR: Sentinel-3 is a near-polar, sun-synchroneous multi-instrument mission whose objective is to measure sea-surface topography, sea- and land-surface temperature, ocean colour and land colour with high-end accuracy and reliability. Sentinel-3A was launched on 16 February 2016 and Sentinel-3B joined its twin in orbit on 25 April 2018.

Following ENVISAT’s AATSR instrument, the SLSTR instrument on-board Sentinel-3 is a conical scanning imaging radiometer employing the along track scanning dual view technique. SLSTR provides data continuity with respect to these previous missions but with a substantial improvement due to its higher swaths. The radiometer has 9 channels ranging from the visible and near-infrared (VNIR) to the thermal infrared (TIR). In single-view mode, SLSTR has a spatial resolution of 1 km, with less than half a day revisit time (with Sentinel-3 A and B) and a swath of 1400 km. In dual-view mode, a 500m resolution is reached, but only covering a swath of 744 Kilometers and with a daily revisit time.

Source of the data: (ESA Copernicus Open Access hub).
Source of the data: for Sentinel-3 Marine and Atmosphere Products,  available through the Copernicus Online Data Access (CODA):

MeteoSAT/SEVIRI: Subskin Sea Surface Temperature can be derived from the Spinning Enhanced Visible and InfraRed Imager (SEVIRI) on Meteosat Second Generation Satellites (MSG). MSG satellites are geostationary satellites covering the East Atlantic and West Indian Oceans. The retrieval of SST from Meteosat/SEVIRI is managed by EUMETSAT OSI-SAF, which provides sub-skin SST data as aggregated (L3C) hourly products remapped onto a 0.05° regular grid. Both reprocessed (from Meteosat-8 and Meteosat-9) and Near-real time (from MteoSAT-11) data are available.

Source of the data: OSI-SAF Eumetsat:

1.3 In situ SST data

There are limited sources of in situ SST observational data. Most in situ data are provided by ocean buoys. Ship observations supplement the buoy data.

NOAA National Data Buoy Center ( maintains an array of moored buoys since 1967, including their collected data. Most of the data are collected hourly, but there is only a very limited number of active buoys—about 50 in total.

Source of the data:,

NOAA Global Drifter Program (GDP) freely drifting buoys also measure SST on an hourly basis.  The program started in 1979 and currently there are ~1,000 active drifters in the global ocean.

Source of the data:

The CMEMS In-Situ Thematic Assembly Center delivers Global Ocean – near real-time (NRT) and Delayed-Mode in situ quality controlled temperature surface and sub-surface observations. NRT products are hourly updated and distributed by INSTAC within 24-48 hours from acquisition in average. Data are collected mainly through global networks (Argo, OceanSites, GOSUD, EGO) and through the GTS. Data are available from 2010-01-15 to present.

Source of the data:

In situ delayed mode products are updated yearly, integrating the best available version of in situ data for temperature and salinity measurements. These data are collected from main global networks (Argo, GOSUD, OceanSITES, World Ocean Database) completed by European data provided by EUROGOOS regional systems and national system by the regional INS TAC components. They cover the period 1950-01-01 to present.

Source of the data:

2.0  Surface Sea Salinity (SSS)

2.1 Remotely sensed SSS datasets and products

The remotely sensed SSS datasets and products are given below.

Aquarius is a microwave instrument designed to measure the surface salinity of the global oceans. It is a combined active/passive L-band microwave instrument in a 3 beam push broom configuration. The instrument’s footprint is 100 km along the 390 km swath. The Aquarius orbit is sun synchronous with a polar trajectory at an altitude of 657 km, with repeat cycle every 7 days. Aquarius started its scientific mission on 27 August 2011 and failed on 7 June 2015.  The retrieved salinity data are provided by PO.DAAC ( as daily, 8-day, monthly, 3-monthly and annual maps on 1 degree lat/lon grid.

Source of the data:

Soil Moisture Active Passive (SMAP) mission was designed to provide measurements of soil moisture in the top 5 cm layer.  The instrument combines a microwave L-band (1.41 GHz) radiometer and  L-band active radar. The footprint of the instrument is about 40 km along a 1000 km-wide swath. SMAP flies on a sun synchronous polar trajectory at an altitude of 685 km. The ground tracks repeat exactly every 8 days, or 117 orbits. The global coverage is achieved every 2-3 days. Even though originally SMAP was designed to measure soil moisture, the microwave radiance data can be used to retrieve surface ocean salinity and wind speed. The SMAP mission begun its operation in the beginning of April 2015 and ended  on 9 September 2016.  The retrieved ocean surface salinities are provided by Remote Sensing Systems as monthly or 8-day averaged maps on 1/4 degree lat/lon grid until the end of July 2016.

Source of the data:

Soil Moisture Ocean Salinity (SMOS) is a European Space Agency (ESA) mission designed to measure soil moisture and ocean surface salinity. The instrument is the Microwave Imaging Radiometer using Aperture Synthesis (MIRAS) 2D interferometric L-band radiometer. It is flying on a sun-synchronous orbit at an altitude of 758 km with 23-day track repeat cycle. The retrieved salinity data are available from May 2010 until now. The final products are distributed by SMOS Barcelona Expert Center ( as daily gridded maps on 1/4 degree lat/lon grid, or as monthly binned maps on 1 degree lat/lon grid.

Source of the data:

2.2 In Situ SSS observation

In situ SSS observations are limited to buoys, ships of opportunity and scientific sea expeditions. The number of buoys equipped with salinity sensors is even smaller than for SST measuring buoys. SSS observations are mostly limited to tropical mooring buoys, ARGO profiling buoys and a limited number of drifters in the Global Drifter Program (GDP).


The CMEMS In-Situ Thematic Assembly Center delivers Global Ocean – near real-time (NRT) and Delayed-Mode in situ, quality controlled, salinity surface and sub-surface observations. NRT products are hourly updated and distributed by INSTAC within 24-48 hours from acquisition in average. Data are collected mainly through global networks (Argo, OceanSites, GOSUD, EGO) and through the GTS. Data are available from 2010-01-15 to present.

Source of the data:

In Situ delayed mode products are updated yearly, integrating the best available version of in situ data for temperature and salinity measurements. These data are collected from main global networks (Argo, GOSUD, OceanSITES, World Ocean Database) completed by European data provided by EUROGOOS regional systems and national system by the regional INS TAC components. They cover the period 1950-01-01 to present.

Source of the data:

3.0 Ocean Currents

3.1 Numerical Modeling approaches

The ocean surface currents are probably one of the most difficult variable to measure accurately.  In many cases we have to rely on numerical modeling approaches.

HYbrid Coordinate Ocean Model  (HYCOM)  is numerical model of global ocean circulation. It is a primitive equation ocean general circulation model with hybrid  vertical coordinate. The horizontal and vertical grid is variable, the nominal horizontal grid spacing is 1/12 deg. In the vertical, the level spacing increases with depth from 2 m at the surface to 1000 m at the bottom of the global oceans (40 depth levels). The atmospheric forcing for HYCOM is the NAVy Global Environmental Model (NAVGEM 2.0). HYCOM dataset time span is from 1992 to present with daily output frequency.

Source of the data:

Global Ocean reanalysis and Simulation (GLORYS 12V1) is a high-resolution global-ocean reanalysis product. The reanalysis is based on real-time global forecasting system CMEMS (Copernicus Marine Environment Monitoring Service), which utilizes NEMO (The Nucleus for European Modeling of the Ocean) ocean model platform. NEMO is driven at the surface by ECMWF ERA-Interim reanalysis data. A reduced-order Kalman filter is employed to assimilate observational data, namely: along track sea level anomaly, satellite SST, sea ice concentration, and in situ temperature and salinity vertical profiles. The reanalysis product is provided on global 1/12 degree lat/lon grid in horizontal and on 50 levels in vertical (-5500.0 to 0.0 m). All relevant oceanic variables are included (temperature, sea level height, currents, salinity etc.) as daily and monthly averages. The time span of the reanalysis data is from 1993-01-01 to 2018-12-25.

Source of the data :

3.2 Observation-based products

Surface CUrrents from Diagnostic model (SCUD), developed by Maximenko and Hafner (2010), uses remotely sensed data of ocean topography (altimetry) and ocean surface wind (QuikSCAT/ ASCAT scatterometry) to derive near-surface currents. The SCUD product assumes that low-frequency near-surface currents are a combination of geostrophic and wind-driven currents with the latter being linear to the (rotated and scaled) wind vector.  Local coefficients of the model are independent of time and optimized using historical data of the Global Drifter Program, comprising the drifters initially drogued at 15 meters but also including their trajectories after the loss of the drogues. The SCUD provides a set of daily surface velocity maps on 1/4 degree grid starting 1999 till present. Initially produced using QuikSCAT wind products, SCUD coefficients were later adjusted to ASCAT and ERA5 winds.

Simulations based on SCUD velocities successfully reproduced drift of debris from the 2011 tsunami in Japan, ONE Apus container spill of December 2020 and several other disasters and accidents. It was also used to plan and coordinate operational surveys and cleanup activities at sea and on the shorelines.

Reference:; Maximenko, N. and J. Hafner, 2010: SCUD: Surface CUrrents from Diagnostic model, IPRC Technical Note No. 5 (SCUD_manual_02_17.pdf)

Source of the data:,

OSCAR is an observation-based product of surface currents only. The surface currents are calculated from satellite data and a simplified physical model of surface mixing layer. The surface current are comprised of three components: geostrophy, wind driven, and thermal adjustment parts. Geostrophy is calculated from satellite-measured sea level height. Satellite derived surface ocean winds and SST are used in the model calculations.   The results are interpreted as average currents in the top 30 m. The data are produced on global 1/3 degree lat/lon grid with 5-day output frequency starting in 1992 until present.


CMEMS/Globcurrent: This product consists of the zonal and meridional total velocities at 0 and 15 m mapped onto a 1/4 degree regular grid. These total velocity fields are obtained by combining CMEMS satellite Geostrophic Surface Currents and modelled Ekman current at the surface and 15 m depth. Products are available both in Near Real Time (daily update) and in Delayed Time (yearly update). In Near-Real time, 6 hourly product, daily, and monthly mean are available from 2020-01-01 to present. In Delayed-Time, 3 hourly product, daily and monthly means are available from 1993-01-01 to present.

Reference: Rio, M.-H., Mulet, S., and Picot, N. (2014), Beyond GOCE for the ocean circulation estimate: Synergetic use of altimetry, gravimetry, and in situ data provides new insight into geostrophic and Ekman currents, Geophys. Res. Lett., 41, 8918– 8925, doi:10.1002/2014GL061773.

Source of the data: for the NRT products for the delayed-time product

CMEMS multi-obs 3D ocean currents: The Copernicus Marine Environment Monitoring Service distributes two observation-based 3D ocean currents: the Multi Observation Global Ocean ARMOR3D L4 analysis and multi-year reprocessing. It consists of 3D Temperature, Salinity, Heights, Geostrophic Currents and Mixed Layer Depth, available on a 1/4 degree regular grid and on 50 depth levels from the surface down to the bottom. The product includes 4 datasets:

    1. dataset-armor-3d-nrt-weekly, which delivers near-real-time (NRT) weekly data
    2. dataset-armor-3d-nrt-monthly, which delivers near-real-time (NRT) monthly data
    3. dataset-armor-3d-rep-weekly, which delivers multi-year reprocessed (REP) weekly data
    4. dataset-armor-3d-rep-monthly, which delivers multi-year reprocessed (REP) monthly data

Reference: Mulet, S., M.-H. Rio, A. Mignot, S. Guinehut and R. Morrow, 2012: A new estimate of the global 3D geostrophic ocean circulation based on satellite data and in-situ measurements. Deep Sea Research Part II : Topical Studies in Oceanography, 77–80(0):70–81.

Source of the data:

The OMEGA3D is observation-based quasi-geostrophic vertical and horizontal ocean currents, developed by the Consiglio Nazionale delle Ricerche (CNR). The data are provided weekly over a regular grid at 1/4° horizontal resolution, from the surface to 1500 m depth (representative of each Wednesday). The velocities are obtained by solving a diabatic formulation of the Omega equation, starting from ARMOR3D data (MULTIOBS_GLO_PHY_REP_015_002) and ERA-Interim surface fluxes.

Reference: Buongiorno Nardelli, B. (2020). CNR global observation-based OMEGA3D quasi-geostrophic vertical and horizontal ocean currents (1993-2018) (Version 1) set. Copernicus Monitoring Environment Marine Service (CMEMS).

Buongiorno Nardelli, B. A Multi-Year Timeseries of Observation-Based 3D Horizontal and Vertical Quasi-Geostrophic Global Ocean Currents. Earth Syst. Sci. Data 2020, No. 12, 1711–1723.

Source of the data:

The World Ocean Circulation (WOC) products: In the frame of the ESA WOC project, whose main objective is to enhance the understanding, the monitoring and the phasing forecast of upper ocean currents, with a particular focus on the most intense ocean small-scale structures, a number of ocean currents products are available for specific regional areas, including Fronts derived from SST observations (Agulhas currents, North Atlantic, Western Europe, 2D merged SSH/SST surface currents over North Atlantic, 3D data-driven currents over North Atlantic, and high resolution 2d surface currents resolving inertial oscillations over North Atlantic.

Source of the data:

High Frequency (HF) Radar measures surface current velocities based on Doppler effect.  The US based program started in 1980’s.  It is land based system located at shorelines. The HF Radar range is limited to 0.3 – 6 km offshore depending on operational frequency, which varies from  4.0  to 45 MHz. Currently  the  HF Radar system covers significant parts of west and east coast of contiguous US, with smaller segments in the Gulf of Mexico and Hawaii.  The data are provided as hourly averages.  The drawback of HF Radar derived currents is its limited spatial coverage to shorelines only.

Source of the data:

3.3 In situ observation – proxy

Global Drifter Program (GDP) buoy data can be viewed as proxy for in situ observations.  GDP buoys are freely drifting buoys with attached drogue at 15 m depth.  The GDP  started operation in 1979,  currently there are over 1,000 GDP drifting buoys deployed in the world’s oceans.  The  location coordinates (longitude/latitude) are collected hourly in real time.   The trajectories  of individual drifting buoys  can be interpreted as average surface currents in  the top 15 m  surface layer.   However, the limitation is in coarse spatial resolution. There are only about 1,000 drifters unevenly distributed in the global oceans.

Source of the data:

The Copernicus In-Situ Thematic Assembly Center (TAC) distributes global ocean-delayed model in-situ observations of surface (drifter and HF Radar) and sub-surface (vessel mounted ADCP) water velocity. This delayed mode product designed for reanalysis purposes integrates the best available version of in situ data for ocean surface currents and current vertical profiles. It concerns three delayed time datasets dedicated to near-surface currents measurements coming from two platforms (Lagrangian surface drifters and High Frequency radars) and velocity profiles within the water column coming from the Acoustic Doppler Current Profiler (ADCP, vessel mounted only) platform

Sources of the data:


Contributing Authors

Jan Hafner, University of Hawaii, USA
Marie-Helene Rio, ESA-ESRIN, Italy
Nikolai Maximenko, University of Hawaii, USA

Start typing and press Enter to search